It’s still early stage so kinda just in a demo state for now. I added a way to configure how long the pause is before it replies, right now it’s at 800 milliseconds just to get a feel for how realtime it is, but I think between 2-4 seconds is probably better for most realistic use cases.

As for the translations, the settings lets you define a support language and a practice language. Basically what you work on is the practice language while the support language is there to help you navigate what’s being said.

I’m currently working on adding some other language learning features, like a grammar book, vocabulary practice and story mode so people train through reading/speaking/writing in other modalities.

Either way it’s going to take a number of iterations and feedback cycles to get the UX right in a way that feels good and allows for progressive growth